ECG Emotion Recognition Using DREAMER Dataset

This project implements a complete MATLAB-based ECG Emotion Recognition System using the DREAMER physiological dataset.
The system extracts ECG signals, preprocesses them, detects R-peaks, computes HRV and amplitude-based features, performs FFT analysis, and classifies emotional states using percentile thresholds.
All generated plots and the final project report are included in this repository.

---

🚀 Overview

• ECG preprocessing (normalization + bandpass filtering)
• R-peak detection using findpeaks()
• RR interval extraction
• HRV and amplitude-based feature analysis
• FFT power spectrum analysis
• Correlation and boxplot visualizations
• Percentile-based emotion classification
• Outputs saved as high-quality plots

---

🫀 ECG Processing Pipeline

1️⃣ Preprocessing

• DC offset removal
• Amplitude normalization
• Bandpass filtering (0.5–50 Hz)

2️⃣ R-Peak Detection

• Detected using height + minimum RR distance constraints
• Used to compute RR intervals and amplitudes

3️⃣ Feature Extraction

• Peak amplitude
• RR intervals
• Moving standard deviation (for dominance)

4️⃣ Frequency-Domain Analysis

• FFT power spectrum shows dominant ECG frequencies
• Useful for stress and arousal estimation

5️⃣ Visualization

Generated charts include:

• Filtered ECG with R-peaks
• RR interval trend
• RR histogram
• FFT power spectrum
• Correlation matrix
• Boxplots
• Stacked emotional sub-feeling distribution

---

😊 Emotion Classification

Emotions are computed using percentile thresholds:

Valence

• Happy
• Sad
• Neutral

Arousal

• Excited
• Calm
• Neutral

Dominance

• Confident
• Passive
• Neutral

These categories are derived using:

• Peak amplitude
• RR interval
• Moving RR standard deviation

---

▶️ How to Run

1️⃣ Download DREAMER dataset

The dataset cannot be included due to licensing.
Download from the official source: https://zenodo.org/record/546113

The script will automatically:

• Load ECG
• Preprocess
• Detect peaks
• Extract features
• Classify emotions
• Generate and save all plots

---

2️⃣ Run the MATLAB Script

matlab ecg_emotion_recognition.m

📘 Full Project Report

📄 Project PDF: ECG_Emotion_Recognition.pdf

---

🛠 Technologies Used

• MATLAB: Signal processing, HRV analysis, FFT
• DREAMER Dataset: Physiological emotion data
• GitHub: Version control & documentation

---

👥 Contributors

• Bojja Divya
• Bhavatarini .M
• Srinithi R.M

---

📜 License

This project is licensed under the MIT License.